src/gromacs/gmxlib/nonbonded/nb_kernel_c/nb_kernel_ElecNone_VdwLJEw_GeomP1P1_c.c

   1 /*
   2  * This file is part of the GROMACS molecular simulation package.
   3  *
   4  * Copyright (c) 2012,2013,2014, by the GROMACS development team, led by
   5  * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
   6  * and including many others, as listed in the AUTHORS file in the
   7  * top-level source directory and at http://www.gromacs.org.
   8  *
   9  * GROMACS is free software; you can redistribute it and/or
  10  * modify it under the terms of the GNU Lesser General Public License
  11  * as published by the Free Software Foundation; either version 2.1
  12  * of the License, or (at your option) any later version.
  13  *
  14  * GROMACS is distributed in the hope that it will be useful,
  15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17  * Lesser General Public License for more details.
  18  *
  19  * You should have received a copy of the GNU Lesser General Public
  20  * License along with GROMACS; if not, see
  21  * http://www.gnu.org/licenses, or write to the Free Software Foundation,
  22  * Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA.
  23  *
  24  * If you want to redistribute modifications to GROMACS, please
  25  * consider that scientific software is very special. Version
  26  * control is crucial - bugs must be traceable. We will be happy to
  27  * consider code for inclusion in the official distribution, but
  28  * derived work must not be called official GROMACS. Details are found
  29  * in the README & COPYING files - if they are missing, get the
  30  * official version at http://www.gromacs.org.
  31  *
  32  * To help us fund GROMACS development, we humbly ask that you cite
  33  * the research papers on the package. Check out http://www.gromacs.org.
  34  */
  35 /*
  36  * Note: this file was generated by the GROMACS c kernel generator.
  37  */
  38 #include "gmxpre.h"
  39
  40 #include "config.h"
  41
  42 #include <math.h>
  43
  44 #include "../nb_kernel.h"
  45 #include "gromacs/legacyheaders/types/simple.h"
  46 #include "gromacs/math/vec.h"
  47 #include "gromacs/legacyheaders/nrnb.h"
  48
  49 /*
  50  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJEw_GeomP1P1_VF_c
  51  * Electrostatics interaction: None
  52  * VdW interaction:            LJEwald
  53  * Geometry:                   Particle-Particle
  54  * Calculate force/pot:        PotentialAndForce
  55  */
  56 void
  57 nb_kernel_ElecNone_VdwLJEw_GeomP1P1_VF_c
  58                     (t_nblist                    * gmx_restrict       nlist,
  59                      rvec                        * gmx_restrict          xx,
  60                      rvec                        * gmx_restrict          ff,
  61                      t_forcerec                  * gmx_restrict          fr,
  62                      t_mdatoms                   * gmx_restrict     mdatoms,
  63                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
  64                      t_nrnb                      * gmx_restrict        nrnb)
  65 {
  66     int              i_shift_offset,i_coord_offset,j_coord_offset;
  67     int              j_index_start,j_index_end;
  68     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
  69     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
  70     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
  71     real             *shiftvec,*fshift,*x,*f;
  72     int              vdwioffset0;
  73     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
  74     int              vdwjidx0;
  75     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
  76     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
  77     int              nvdwtype;
  78     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
  79     int              *vdwtype;
  80     real             *vdwparam;
  81     real             c6grid_00;
  82     real             ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
  83     real             *vdwgridparam;
  84
  85     x                = xx[0];
  86     f                = ff[0];
  87
  88     nri              = nlist->nri;
  89     iinr             = nlist->iinr;
  90     jindex           = nlist->jindex;
  91     jjnr             = nlist->jjnr;
  92     shiftidx         = nlist->shift;
  93     gid              = nlist->gid;
  94     shiftvec         = fr->shift_vec[0];
  95     fshift           = fr->fshift[0];
  96     nvdwtype         = fr->ntype;
  97     vdwparam         = fr->nbfp;
  98     vdwtype          = mdatoms->typeA;
  99     vdwgridparam     = fr->ljpme_c6grid;
 100     ewclj            = fr->ewaldcoeff_lj;
 101     sh_lj_ewald      = fr->ic->sh_lj_ewald;
 102     ewclj2           = ewclj*ewclj;
 103     ewclj6           = ewclj2*ewclj2*ewclj2;
 104
 105     outeriter        = 0;
 106     inneriter        = 0;
 107
 108     /* Start outer loop over neighborlists */
 109     for(iidx=0; iidx<nri; iidx++)
 110     {
 111         /* Load shift vector for this list */
 112         i_shift_offset   = DIM*shiftidx[iidx];
 113         shX              = shiftvec[i_shift_offset+XX];
 114         shY              = shiftvec[i_shift_offset+YY];
 115         shZ              = shiftvec[i_shift_offset+ZZ];
 116
 117         /* Load limits for loop over neighbors */
 118         j_index_start    = jindex[iidx];
 119         j_index_end      = jindex[iidx+1];
 120
 121         /* Get outer coordinate index */
 122         inr              = iinr[iidx];
 123         i_coord_offset   = DIM*inr;
 124
 125         /* Load i particle coords and add shift vector */
 126         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 127         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 128         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 129
 130         fix0             = 0.0;
 131         fiy0             = 0.0;
 132         fiz0             = 0.0;
 133
 134         /* Load parameters for i particles */
 135         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 136
 137         /* Reset potential sums */
 138         vvdwsum          = 0.0;
 139
 140         /* Start inner kernel loop */
 141         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 142         {
 143             /* Get j neighbor index, and coordinate index */
 144             jnr              = jjnr[jidx];
 145             j_coord_offset   = DIM*jnr;
 146
 147             /* load j atom coordinates */
 148             jx0              = x[j_coord_offset+DIM*0+XX];
 149             jy0              = x[j_coord_offset+DIM*0+YY];
 150             jz0              = x[j_coord_offset+DIM*0+ZZ];
 151
 152             /* Calculate displacement vector */
 153             dx00             = ix0 - jx0;
 154             dy00             = iy0 - jy0;
 155             dz00             = iz0 - jz0;
 156
 157             /* Calculate squared distance and things based on it */
 158             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 159
 160             rinv00           = gmx_invsqrt(rsq00);
 161
 162             rinvsq00         = rinv00*rinv00;
 163
 164             /* Load parameters for j particles */
 165             vdwjidx0         = 2*vdwtype[jnr+0];
 166
 167             /**************************
 168              * CALCULATE INTERACTIONS *
 169              **************************/
 170
 171             r00              = rsq00*rinv00;
 172
 173             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 174             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 175             c6grid_00        = vdwgridparam[vdwioffset0+vdwjidx0];
 176
 177             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 178             ewcljrsq         = ewclj2*rsq00;
 179             exponent         = exp(-ewcljrsq);
 180             poly             = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
 181             vvdw6            = (c6_00-c6grid_00*(1.0-poly))*rinvsix;
 182             vvdw12           = c12_00*rinvsix*rinvsix;
 183             vvdw             = vvdw12*(1.0/12.0) - vvdw6*(1.0/6.0);
 184             fvdw             = (vvdw12 - vvdw6 - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
 185
 186             /* Update potential sums from outer loop */
 187             vvdwsum         += vvdw;
 188
 189             fscal            = fvdw;
 190
 191             /* Calculate temporary vectorial force */
 192             tx               = fscal*dx00;
 193             ty               = fscal*dy00;
 194             tz               = fscal*dz00;
 195
 196             /* Update vectorial force */
 197             fix0            += tx;
 198             fiy0            += ty;
 199             fiz0            += tz;
 200             f[j_coord_offset+DIM*0+XX] -= tx;
 201             f[j_coord_offset+DIM*0+YY] -= ty;
 202             f[j_coord_offset+DIM*0+ZZ] -= tz;
 203
 204             /* Inner loop uses 49 flops */
 205         }
 206         /* End of innermost loop */
 207
 208         tx = ty = tz = 0;
 209         f[i_coord_offset+DIM*0+XX] += fix0;
 210         f[i_coord_offset+DIM*0+YY] += fiy0;
 211         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 212         tx                         += fix0;
 213         ty                         += fiy0;
 214         tz                         += fiz0;
 215         fshift[i_shift_offset+XX]  += tx;
 216         fshift[i_shift_offset+YY]  += ty;
 217         fshift[i_shift_offset+ZZ]  += tz;
 218
 219         ggid                        = gid[iidx];
 220         /* Update potential energies */
 221         kernel_data->energygrp_vdw[ggid] += vvdwsum;
 222
 223         /* Increment number of inner iterations */
 224         inneriter                  += j_index_end - j_index_start;
 225
 226         /* Outer loop uses 13 flops */
 227     }
 228
 229     /* Increment number of outer iterations */
 230     outeriter        += nri;
 231
 232     /* Update outer/inner flops */
 233
 234     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_VF,outeriter*13 + inneriter*49);
 235 }
 236 /*
 237  * Gromacs nonbonded kernel:   nb_kernel_ElecNone_VdwLJEw_GeomP1P1_F_c
 238  * Electrostatics interaction: None
 239  * VdW interaction:            LJEwald
 240  * Geometry:                   Particle-Particle
 241  * Calculate force/pot:        Force
 242  */
 243 void
 244 nb_kernel_ElecNone_VdwLJEw_GeomP1P1_F_c
 245                     (t_nblist                    * gmx_restrict       nlist,
 246                      rvec                        * gmx_restrict          xx,
 247                      rvec                        * gmx_restrict          ff,
 248                      t_forcerec                  * gmx_restrict          fr,
 249                      t_mdatoms                   * gmx_restrict     mdatoms,
 250                      nb_kernel_data_t gmx_unused * gmx_restrict kernel_data,
 251                      t_nrnb                      * gmx_restrict        nrnb)
 252 {
 253     int              i_shift_offset,i_coord_offset,j_coord_offset;
 254     int              j_index_start,j_index_end;
 255     int              nri,inr,ggid,iidx,jidx,jnr,outeriter,inneriter;
 256     real             shX,shY,shZ,tx,ty,tz,fscal,rcutoff,rcutoff2;
 257     int              *iinr,*jindex,*jjnr,*shiftidx,*gid;
 258     real             *shiftvec,*fshift,*x,*f;
 259     int              vdwioffset0;
 260     real             ix0,iy0,iz0,fix0,fiy0,fiz0,iq0,isai0;
 261     int              vdwjidx0;
 262     real             jx0,jy0,jz0,fjx0,fjy0,fjz0,jq0,isaj0;
 263     real             dx00,dy00,dz00,rsq00,rinv00,rinvsq00,r00,qq00,c6_00,c12_00,cexp1_00,cexp2_00;
 264     int              nvdwtype;
 265     real             rinvsix,rvdw,vvdw,vvdw6,vvdw12,fvdw,fvdw6,fvdw12,vvdwsum,br,vvdwexp,sh_vdw_invrcut6;
 266     int              *vdwtype;
 267     real             *vdwparam;
 268     real             c6grid_00;
 269     real             ewclj,ewclj2,ewclj6,ewcljrsq,poly,exponent,sh_lj_ewald;
 270     real             *vdwgridparam;
 271
 272     x                = xx[0];
 273     f                = ff[0];
 274
 275     nri              = nlist->nri;
 276     iinr             = nlist->iinr;
 277     jindex           = nlist->jindex;
 278     jjnr             = nlist->jjnr;
 279     shiftidx         = nlist->shift;
 280     gid              = nlist->gid;
 281     shiftvec         = fr->shift_vec[0];
 282     fshift           = fr->fshift[0];
 283     nvdwtype         = fr->ntype;
 284     vdwparam         = fr->nbfp;
 285     vdwtype          = mdatoms->typeA;
 286     vdwgridparam     = fr->ljpme_c6grid;
 287     ewclj            = fr->ewaldcoeff_lj;
 288     sh_lj_ewald      = fr->ic->sh_lj_ewald;
 289     ewclj2           = ewclj*ewclj;
 290     ewclj6           = ewclj2*ewclj2*ewclj2;
 291
 292     outeriter        = 0;
 293     inneriter        = 0;
 294
 295     /* Start outer loop over neighborlists */
 296     for(iidx=0; iidx<nri; iidx++)
 297     {
 298         /* Load shift vector for this list */
 299         i_shift_offset   = DIM*shiftidx[iidx];
 300         shX              = shiftvec[i_shift_offset+XX];
 301         shY              = shiftvec[i_shift_offset+YY];
 302         shZ              = shiftvec[i_shift_offset+ZZ];
 303
 304         /* Load limits for loop over neighbors */
 305         j_index_start    = jindex[iidx];
 306         j_index_end      = jindex[iidx+1];
 307
 308         /* Get outer coordinate index */
 309         inr              = iinr[iidx];
 310         i_coord_offset   = DIM*inr;
 311
 312         /* Load i particle coords and add shift vector */
 313         ix0              = shX + x[i_coord_offset+DIM*0+XX];
 314         iy0              = shY + x[i_coord_offset+DIM*0+YY];
 315         iz0              = shZ + x[i_coord_offset+DIM*0+ZZ];
 316
 317         fix0             = 0.0;
 318         fiy0             = 0.0;
 319         fiz0             = 0.0;
 320
 321         /* Load parameters for i particles */
 322         vdwioffset0      = 2*nvdwtype*vdwtype[inr+0];
 323
 324         /* Start inner kernel loop */
 325         for(jidx=j_index_start; jidx<j_index_end; jidx++)
 326         {
 327             /* Get j neighbor index, and coordinate index */
 328             jnr              = jjnr[jidx];
 329             j_coord_offset   = DIM*jnr;
 330
 331             /* load j atom coordinates */
 332             jx0              = x[j_coord_offset+DIM*0+XX];
 333             jy0              = x[j_coord_offset+DIM*0+YY];
 334             jz0              = x[j_coord_offset+DIM*0+ZZ];
 335
 336             /* Calculate displacement vector */
 337             dx00             = ix0 - jx0;
 338             dy00             = iy0 - jy0;
 339             dz00             = iz0 - jz0;
 340
 341             /* Calculate squared distance and things based on it */
 342             rsq00            = dx00*dx00+dy00*dy00+dz00*dz00;
 343
 344             rinv00           = gmx_invsqrt(rsq00);
 345
 346             rinvsq00         = rinv00*rinv00;
 347
 348             /* Load parameters for j particles */
 349             vdwjidx0         = 2*vdwtype[jnr+0];
 350
 351             /**************************
 352              * CALCULATE INTERACTIONS *
 353              **************************/
 354
 355             r00              = rsq00*rinv00;
 356
 357             c6_00            = vdwparam[vdwioffset0+vdwjidx0];
 358             c12_00           = vdwparam[vdwioffset0+vdwjidx0+1];
 359             c6grid_00        = vdwgridparam[vdwioffset0+vdwjidx0];
 360
 361             rinvsix          = rinvsq00*rinvsq00*rinvsq00;
 362             ewcljrsq         = ewclj2*rsq00;
 363             exponent         = exp(-ewcljrsq);
 364             poly             = exponent*(1.0 + ewcljrsq + ewcljrsq*ewcljrsq*0.5);
 365             fvdw             = (((c12_00*rinvsix - c6_00 + c6grid_00*(1.0-poly))*rinvsix) - c6grid_00*(1.0/6.0)*exponent*ewclj6)*rinvsq00;
 366
 367             fscal            = fvdw;
 368
 369             /* Calculate temporary vectorial force */
 370             tx               = fscal*dx00;
 371             ty               = fscal*dy00;
 372             tz               = fscal*dz00;
 373
 374             /* Update vectorial force */
 375             fix0            += tx;
 376             fiy0            += ty;
 377             fiz0            += tz;
 378             f[j_coord_offset+DIM*0+XX] -= tx;
 379             f[j_coord_offset+DIM*0+YY] -= ty;
 380             f[j_coord_offset+DIM*0+ZZ] -= tz;
 381
 382             /* Inner loop uses 44 flops */
 383         }
 384         /* End of innermost loop */
 385
 386         tx = ty = tz = 0;
 387         f[i_coord_offset+DIM*0+XX] += fix0;
 388         f[i_coord_offset+DIM*0+YY] += fiy0;
 389         f[i_coord_offset+DIM*0+ZZ] += fiz0;
 390         tx                         += fix0;
 391         ty                         += fiy0;
 392         tz                         += fiz0;
 393         fshift[i_shift_offset+XX]  += tx;
 394         fshift[i_shift_offset+YY]  += ty;
 395         fshift[i_shift_offset+ZZ]  += tz;
 396
 397         /* Increment number of inner iterations */
 398         inneriter                  += j_index_end - j_index_start;
 399
 400         /* Outer loop uses 12 flops */
 401     }
 402
 403     /* Increment number of outer iterations */
 404     outeriter        += nri;
 405
 406     /* Update outer/inner flops */
 407
 408     inc_nrnb(nrnb,eNR_NBKERNEL_VDW_F,outeriter*12 + inneriter*44);
 409 }